Volume 19, Issue 3, Pages 547-556 (March 2011) Notable Reduction in Illegitimate Integration Mediated by a PPT-deleted, Nonintegrating Lentiviral Vector  Boris Kantor, Matthew Bayer, Hong Ma, Jude Samulski, Chengwen Li, Thomas McCown, Tal Kafri  Molecular Therapy  Volume 19, Issue 3, Pages 547-556 (March 2011) DOI: 10.1038/mt.2010.277 Copyright © 2011 The American Society of Gene & Cell Therapy Terms and Conditions

Figure 1 Analysis of lentiviral vector episome formation under conditions not conducive to homologous recombination. Shuttle vector analysis (see Materials and Methods) was employed on Hirt-extracted episomes from lentiviral shuttle vector-transduced (a) Ercc1-defective Chinese hamster ovary (E1KO7-5) cells, (b) liver cells, and (c) cycle-arrested Chinese hamster ovary (AA8 cells). Panels a–c represent episome formation as a percentage of total episomes; each experiment was performed in triplicate. The term “mutant” refers to circular vector genomes generated by autointegration, as determined by restriction products that do not fit the expected band sizes corresponding to 2-long terminal repeat (LTR) and 1-LTR circular episomes. (d) Episome formation was examined by Southern-blot analysis on DNA extracted 2 days after transduction with an integrating, polypurine tract+ vector (vTK945) from Ercc1-deficient (E1K07-5) (lane 1), Ercc1-wild-type (ATStg) (lane 2), cycling AA8 (lane 3), arrested AA8 (lane 4), human breast cancer BRCA1-deficient (SUM149) cells (lane 5), and human breast cancer (ME16C2) cells containing wild-type BRCA1 (lane 6). DNA was digested with EcoNI and PflMI and probed with the probe complementary to the 3′ region of vTK945, as shown in Figure 3b. Molecular Therapy 2011 19, 547-556DOI: (10.1038/mt.2010.277) Copyright © 2011 The American Society of Gene & Cell Therapy Terms and Conditions

Figure 2 Proposed model for episome formation and integration in cells transduced with polypurine tract (PPT)-positive or PPT-deleted vectors. (a) The RNA genomes of PPT-positive vectors undergo plus-strand synthesis (depicted as a dashed line) initiation at the PPT, leading to efficient strand displacement and producing primarily linear episomes, with a small number of 1-long terminal repeat (LTR) circles formed by failed strand displacement. Linear episomes may then serve as the substrates for homologous recombination (HR), forming 1-LTR circles; nonhomologous end-joining (NHEJ), forming 2-LTR circles; and integrase-mediated integration, forming integrated provirus. (b) The RNA genomes of PPT-deleted (ΔPPT) vectors undergo plus-strand synthesis initiation at cryptic sites upstream or downstream of the PPT, leading to inefficient strand displacement and producing primarily 1-LTR circles (plus-strand synthesis is represented by dashed lines). However, in the few instances in which initiation of plus-strand synthesis from non-PPT sites (from sites located either upstream to the deleted PPT or in the 3′U3) is followed by strand displacement, the aberrant episomal linear forms contained either additions to their 5′ ends or deletions in the 5′ U3 region. NHEJ-mediated circularization of these aberrant linear forms generate 2-LTR circles containing either an insertion between the LTRs or a deletion in the 5′ U3 region. The aberrant linear forms that lack the 5′ integrase attachment (att) site do not support efficient integrase-mediated integration. Molecular Therapy 2011 19, 547-556DOI: (10.1038/mt.2010.277) Copyright © 2011 The American Society of Gene & Cell Therapy Terms and Conditions

Figure 3 Analyzing the effect of the polypurine tract (PPT) deletion on lentiviral vector episome formation. (a,b) Outline of Southern-blot analysis used to characterize vector provirus and episomes by probing at the (a) 5′ and (b) 3′ regions of the genome. After extracting vector DNA from transduced cells, digestion with EcoNI and PflMI, and probing with either (a) a KasI/BamHI fragment of at PPT+ vector (vTK945) or (b) a AfeI/EcoRV fragment of vTK945, bands indicative of total episomes and integrated provirus (2.7 kb), 2-long terminal repeat (LTR) circles (1.6 kb), 1-LTR circles (1.3 kb), and linear episomes (0.8 kb) were determined. (c) Cells were transduced with a PPT+ vector (vTK945), integrating (lanes 1 and 5) or nonintegrating (lanes 3 and 7), or with a PPT− vector (vTK1179), integrating (lanes 2 and 6) or nonintegrating (lanes 4 and 8) and total DNA was extracted from transduced cells 3 days (no passages) (lanes 1–4) or ~14 days (four passages (p4)) (lanes 5–8) post-transduction, digested as shown in a and b, and analyzed by Southern blot, using the EcoNI-spanning probe complementary to the 5′ region of the virus. (d) Cells were transduced with the PPT+ vector, integrating (lanes 1 and 5) or nonintegrating (lanes 2 and 6), or with the PPT− vector, integrating (lanes 3 and 7) or nonintegrating (lanes 4 and 8) and total DNA was extracted from transduced cells 3 days (no passages) (lanes 1–4) or ~14 days (p4) (lanes 5–8) post-transduction, digested as in (a,b), and analyzed by Southern blot, using the PflMI-spanning probe complementary to the 3′ region of the virus. For c,d, bands corresponding to total vector genomes, 2-LTR circular, 1-LTR circular, and linear episomal genomes, as outlined in a,b, respectively, are shown. (e) Sixteen hours after transduction with a non-self-inactivating (non-SIN), PPT-deleted shuttle vector (vTK1074) and its PPT-positive equivalent (vTK459), which were packaged with both functional integrase (IN+) and mutant integrase (IN−), episomes were harvested from 293T cells and analyzed by shuttle–vector assay; each transduction was performed in triplicate. (f) Episomes were harvested from 293T cells 16 hours after transduction with a SIN, PPT-deleted shuttle vector (vTK1046) and its PPT-positive equivalent (vTK1054), which were packaged with IN+ and IN−, and analyzed by shuttle–vector assay; each transduction was performed in triplicate. For e,f, the data are presented as percentages of 1-LTR, 2-LTR, and mutant forms out of the number of total episomes. Molecular Therapy 2011 19, 547-556DOI: (10.1038/mt.2010.277) Copyright © 2011 The American Society of Gene & Cell Therapy Terms and Conditions

Figure 4 Analyzing the effect of the polypurine tract (PPT) deletion on in vitro expression and integrase-mediated or integrase-independent integration from lentiviral vectors. (a,b) Fluorescence-activated cell sorting analysis of green fluorescent protein (GFP) expression generated by PPT+ (vTK1187) (left) or PPT− (vTK1188) (right), with (IN+) (upper) or without (lower) functional integrase (IN−), in 293T cells was carried out either (a) at 3 days (passage 0 (p0)) or (b) four passages (p4) post-transduction. The percentage of GFP-positive cells and mean fluorescence intensity (MFI) are shown. (c) Quantification in 293T cells of vector integration by PPT+ (vTK1187) (dark gray) or PPT− (vTK1188) (light gray), with (IN+) (left side) or without (IN−) (right side) functional integrase. Integration percentage was calculated for each vector as a ratio of vector copy number per cell at p4 and p0. Samples subjected to the qPCR assay were analyzed in triplicate, and error bars are presented as ± SD. Molecular Therapy 2011 19, 547-556DOI: (10.1038/mt.2010.277) Copyright © 2011 The American Society of Gene & Cell Therapy Terms and Conditions

Figure 5 The effect of the polypurine tract (PPT) deletion on transgene expression from lentiviral vectors in vivo. Green fluorescent protein expression mediated by PPT+ (vTK945) or PPT− (vTK1023) vectors, packaged into particles containing either functional (IN+) (upper) or deficient (IN−) (lower) integrase, in the striatum of the rat brain, was measured 2 months after vector infusion. The indicated scale bar is equal to 20 µm. Molecular Therapy 2011 19, 547-556DOI: (10.1038/mt.2010.277) Copyright © 2011 The American Society of Gene & Cell Therapy Terms and Conditions